A merged lung cancer transcriptome dataset for clinical predictive modeling

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• The Gene Expression Omnibus (GEO) database offers vast transcriptomic data.
• Analyzing large-scale genomic data requires specialized bioinformatics skills, limiting accessibility.
• Challenges exist in data analysis, sharing, and visualization for non-experts.

Purpose of the Study:

• To develop an integrated bioinformatics pipeline for accessible transcriptomic data analysis.
• To create a normalized, preprocessed dataset for non-small cell lung cancer (NSCLC) meta-analysis.
• To facilitate biomarker discovery by simplifying data mining and statistical correction.

Main Methods:

• Integrated multiple open-source R packages into a bioinformatics workflow.
• Merged, normalized, and batch-corrected data from ten independent GEO datasets.
• Filtered genes with low variance and incorporated clinical metadata.

Main Results:

• Developed a normalized dataset comprising 1,118 patient-derived NSCLC samples and paired normal lung tissues.
• The dataset is preprocessed using robust statistical methods for differential expression analysis.
• The workflow enables large-scale meta-analysis without extensive data mining.

Conclusions:

• The integrated pipeline and normalized dataset enhance accessibility of large-scale genomic data.
• This resource serves as a powerful tool for early predictive biomarker discovery in NSCLC.
• The study overcomes technical barriers in transcriptomic data analysis for broader research application.